By KEY TAKEAWAYS
- You can measure AI visibility for free: run a fixed set of 20–30 buyer-style prompts through each AI engine on a schedule and log whether you are mentioned, cited and recommended.
- Track mentions, cited URLs and share of voice against named competitors separately for each engine, because ChatGPT, Google AI Overviews and AI Mode often name different brands for the same question.
- Single snapshots mislead because AI answers change between runs, so repeat each prompt several times and report trends, not one-off screenshots.
You can’t improve what you can’t see, and right now most brands can’t see how they appear when someone asks ChatGPT, Perplexity or Google’s AI Mode for a recommendation.
Classic rank trackers don’t help much here. There is no “position 3” in a conversational answer, and the answer can change the next time you ask. So how do you measure it?
This guide gives you a free method you can start today, the metrics worth reporting, a look at the tool landscape, and an honest list of what AI visibility tracking can’t tell you.
A note on sources: most published guidance on this topic comes from companies that sell tracking tools. We treat their numbers as directional, and we say so where it matters.
What does “AI visibility” actually mean?
AI visibility is how often, and how favorably, your brand appears in AI-generated answers to the questions your buyers ask. It has three parts: being mentioned by name, having your URL cited as a source, and being recommended over alternatives. You can have one without the others.
That last point matters. Tracking vendors describe “ghost citations,” where an AI cites your page without naming your brand. If you only search for your brand name, you miss those. Track both mentions and cited URLs.
Which metrics should you track?
Track five numbers per engine: visibility rate, citation rate, average position, share of voice and sentiment. These are the definitions most tracking guides converge on (Searchable).
| Metric | What it measures | How to calculate it |
|---|---|---|
| Visibility rate | How often you are mentioned | Runs with a mention ÷ total runs |
| Citation rate | How often your URL is linked | Runs citing your domain ÷ total runs |
| Average position | Where you appear in the answer | Order of brands named, counted from the top |
| Share of voice | Your slice versus competitors | Your mentions ÷ all tracked-brand mentions |
| Sentiment | Tone when you are named | Positive, neutral or negative, tagged by hand |
Keep it simple at first. If you only track two, make them visibility rate and share of voice.
How do you build a prompt set?
Start with 20–30 prompts that mirror real buyer questions across the funnel. Your prompt set is your measuring stick, so spend an hour on it.
Pull prompts from these places:
- Sales calls and discovery-call notes (“what did the buyer ask before they found us?”)
- Google Search Console queries phrased as questions
- Your support inbox and live-chat transcripts
- People Also Ask boxes for your core topics
Then sort them into three buckets:
- Discovery: “What are the best [category] tools for [use case]?”
- Comparison: “[Your brand] vs [competitor]: which is better for [situation]?”
- Decision: “Is [your brand] worth it?” or “What does [service] cost?”
Write them the way people actually type or speak. Long, specific, conversational prompts beat short keywords.
How do you track AI visibility for free?
Run your prompt set on a schedule in each engine and log the results in a spreadsheet. It takes about 90 minutes a week for 30 prompts across three engines, and it is enough to establish a baseline.
Here is the method we use as a starting point with clients:
- Set up a sheet with one row per prompt per engine per run. Columns: date, engine, prompt, mentioned (Y/N), cited URL (Y/N), position, competitors named, sentiment, notes.
- Use a clean session. Log out, use a private window, or use a fresh chat so your history doesn’t bias results.
- Run each prompt three times per session window. AI answers change between runs, and a single snapshot misleads. One source says you’d need hundreds of runs for a number you can fully trust, which is why paid tools exist (Searchable). For a free baseline, three runs is a practical compromise.
- Track each engine separately: ChatGPT, Perplexity, Google AI Overviews and AI Mode, and Gemini or Copilot if your audience uses them. A brand can dominate Perplexity citations and be absent from ChatGPT.
- Log competitors every time. Share of voice needs a named competitive set of 4–6 brands.
- Review monthly. Compare averages, not individual runs.
Want a free shortcut? Ahrefs offers a free AI Visibility Checker covering several AI engines with no signup, per its page (Ahrefs). Use it as a cross-check on your manual numbers, not a replacement.
How different are the engines, really?
Quite different. Tracking guides cite research comparing Google AI Overviews, AI Mode and ChatGPT on identical queries that found the three named the same brands only about a third of the time (Searchable). Treat that as a vendor-reported figure, but the direction matches what we see: don’t assume a win in one engine carries over to another.
Our read: this is why a single “AI visibility score” can hide the real story. Always break results out by engine.
When are paid tools worth it?
Paid tools earn their place when manual tracking stops scaling. That usually means daily runs, hundreds of prompts, several countries, or client-ready reporting.
The market is crowded and moves quickly, so check current plans before buying. Examples that appeared in recent roundups include Peec AI and LLMrefs, with entry plans listed at roughly $95 and $79 a month respectively at the time of the sources we reviewed. Verify before you rely on those prices.
Use this checklist when comparing tools:
- Which engines does it cover, and does it include Google AI Mode?
- Does it run each prompt multiple times and show averages?
- Can you set your own prompts, not just its suggestions?
- Does it track cited URLs as well as brand mentions?
- Can you export raw answers so you can audit them?
- Does the price scale sensibly with prompts and countries?
Can an AI agent do the tracking for you?
Yes, with caveats. You can wire an automation to run your prompts through each engine’s API, extract brand mentions and sources, and append rows to your sheet. We cover the building blocks in our guide to AI agents for marketing teams.
Two cautions. API answers can differ from what users see in the consumer apps, and some engines restrict automated querying in their terms. Read the terms and spot-check against manual runs before you trust the output.
What can’t AI visibility tracking tell you?
It can’t prove revenue impact on its own, and it can’t tell you why an answer changed. Be upfront about the limits:
- Answers vary by user, location, time and model version.
- Prompt sets are a sample, not the whole market.
- Mentions don’t equal clicks, and many AI answers produce no click at all.
- Vendor benchmarks are marketing-adjacent. Prefer your own data.
Pair the tracking with referral traffic from AI engines in your analytics and with “how did you hear about us?” questions on forms.
What do you do with the results?
Turn gaps into a content and authority to-do list. If a competitor owns a prompt, open the cited sources and see what they have that you don’t: a comparison table, fresher data, clearer definitions or a third-party review.
Our GEO guide explains how to fix those gaps, and SEO vs GEO vs AEO shows how to run both programs without doubling your workload. Our free AI Search Checklist gives you the page-level fixes.
What to do this week
- Write your prompt set: 20–30 real buyer questions, sorted into discovery, comparison and decision.
- Name your competitive set: pick 4–6 brands to track in every run.
- Build the spreadsheet with the columns above and run a first baseline in ChatGPT, Perplexity and Google AI Mode.
- Cross-check one engine with a free tool and note any differences.
- Pick one gap: choose the prompt where you matter most and a rival wins, and plan the page that would earn the citation.
Frequently asked questions
How do I check if my brand shows up in ChatGPT?
+
Write 20 to 30 questions your buyers really ask, run each one in ChatGPT in a logged-out or fresh session, and record whether your brand is named, whether your URL is cited, and which competitors appear. Repeat weekly. One run proves little, because answers vary between sessions and over time.
What is AI share of voice?
+
AI share of voice is how often your brand is named in AI-generated answers compared with competitors, across a fixed set of prompts. If you and four rivals produce 100 brand mentions and you account for 20, your share of voice is 20 percent. Calculate it per engine, not as one blended number.
Do I need a paid AI visibility tool?
+
Not to start. A spreadsheet and 30 prompts will tell you where you stand. Paid tools earn their place when you need daily runs, many prompts, several countries or reporting for clients, because manual tracking gets slow. Vendor pricing changes often, so check current plans before you buy.
How often should I track AI search visibility?
+
Weekly is enough for most small teams, with a fuller monthly review. AI answers change between runs, so a single check is noise. Run each prompt a few times per session window and compare averages over several weeks before you conclude that anything improved or declined.
Why does my brand appear in Perplexity but not ChatGPT?
+
Each engine retrieves and ranks sources differently, so visibility rarely matches across platforms. Research summarized by tracking vendors found that Google AI Overviews, AI Mode and ChatGPT named the same brands only about a third of the time. Track each engine separately and fix gaps with better, more citable pages.
Sources
WRITTEN BY
Junaid Mubashar
Founder of an AI-native growth agency. Junaid plans and runs full-funnel marketing (SEO and AI search, paid social, Google Ads) and builds the AI workflows and agents that automate it. Everything on this blog is tested with real campaigns before we recommend it.
🔍 The AI Search (GEO) Checklist
42 checks to get your brand cited by ChatGPT, Perplexity, Gemini and Google AI Overviews.